model-validation-and-uncertainty

What problem does it solve?

Evaluates and quantifies the performance and reliability of predictive ecological models by computing standard metrics, evaluating calibration, and mapping uncertainty across predictions.

Core Features & Use Cases

• Comprehensive performance metrics: AUC-ROC, TSS, calibration curves, and threshold analysis for SDMs, occupancy models, and ensemble projections.
• External validation & transferability: supports application to independent datasets or different time periods to assess generalization.
• Uncertainty quantification & visualization: provides bootstrap/ensemble uncertainty maps and a narrative validation report.

Quick Start

Provide a fitted model object and an independent validation dataset to run the evaluation and generate the diagnostic outputs.